khalas
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local binary named
khalasto perform token analysis and optimization tasks through shell commands. - [PROMPT_INJECTION]: The skill contains a surface for indirect prompt injection by processing untrusted user data in shell arguments.
- Ingestion points: User-provided text is passed directly as arguments to the
khalastool inSKILL.md. - Boundary markers: The documentation uses double quotes for input wrapping, which is a weak boundary against sophisticated shell injection.
- Capability inventory: Performs shell execution with variable user-controlled input.
- Sanitization: No input validation, escaping, or filtering is specified for the text before it is passed to the command line.
Audit Metadata